With Linux 3.18-rc1 having came one week early, the EXT4 file-system pull request didn't end up landing until today. However, the EXT4 changes aren't overly exciting for the 3.18 merge window.
The EXT4 changes for the Linux 3.18 kernel merge window come down to mostly just code clean-ups and bug fixes along with some minor journal optimizations.
The NVPTX back-end code for GCC that's going to allow OpenACC 2.0 offloading support for NVIDIA GPUs with GCC is close to materializing within the mainline code-base.
For the past year Mentor Graphics / Code Sourcery has been working on OpenACC 2.0 with GPU offloading as a big addition to the GNU Compiler Collection through their work with NVIDIA Corp. The offloading infrastructure has been worked on for a while and the code that soon looks like it will land is the NVPTX support.
For now it looks like the Linux kernel is going to explicitly declare itself as using the GNU89 dialect of the C89 standard but over time the code is being made to compile under C11.
With GCC 5 switching to use GNU11 by default over the GNU89 C standard (and Clang also moving to GNU11 by default), a patch was proposed on the Linux kernel mailing list this weekend for explicitly using GNU89 as up to now the -std=gnu89 compiler switch wasn't set by default so if using GCC 5 it will try to compile the kernel against GNU11.
Synopsis announced an “HS38″ version of its Linux-focused DesignWare ARC core IP with a new ARCv2 ISA and support for 2.2GHz, 4200 DMIPS speeds at 28nm.
Synopsis acquired its Linux-optimized line of DesignWare ARC 32-bit RISC/DSP cores when it bought semiconductor IP vendor Virage Logic back in 2010 shortly after Virage acquired ARC International. Since then Synopsis has released several DesignWare ARC HS processor designs, most recently with the HS36.
Companies are increasingly turning to cloud services to build and deliver their applications, but those that want to use an open source cloud may find it more difficult to set up and maintain. Service-providers such as UK-based DataCentred can more efficiently set up an enterprise cloud using open source software, at scale.
Hosting provider OVH has launched a cloud service based on IBM’s Power8 processor architecture, an open source architecture tailored specifically for big data applications, and OpenStack.
OVH, which serves 700,000 customers from 17 datacentres globally, said it wanted to provide a robust public cloud service tuned for database workloads and has tapped a combination of IBM and OpenStack-based technologies in this pursuit.
Tails 1.2 is released and announced by Tails developers bring with new feature and improvement. As we know, Tails is a live linux distribution based on debian and focused to preserve your privacy and anonymity. It helps you to use the Internet anonymously and circumvent censorship almost anywhere you go and on any computer but leaving no trace unless you ask it to explicitly.
One of the most important changes in this release is tails 1.2 not included Iceweasel Internet browser and it replaced with Tor Browser, which is based on the latest 4.0 release and Firefox 31.2 ESR. Also, all the applications of tails 1.2 now confined with Apparmor. The Linux kernel has been updated to version 3.16.5-1, and VirtualBox guest additions should now work by default, improving the performance of the OS in a virtual machine.
So when I released 3.17, I said that I'd extend the merge window to
three weeks due to travel.
I clearly lied.
Because here we are, the usual two weeks later, and I've already
pushed out 3.18-rc1.
What happened is that not only did I merge actively despite travels -
I was out of communication just for a couple of days (almost, but not
entirely, due to flights - the hotel in DÃsseldorf lost all internet
for a day too). But perhaps more importantly, people seem to have
aggressively sent in their pull requests, because rc1 contains more
than linux-next did a couple of days after 3.17.. So holding it up
another week just seems pointless.
That said, I realize that people might have taken my statements at
face value, and planned with that in mind. I hate it when I get pull
requests really late in the merge window, but having closed it as per
the regular schedule, I also understand that somebody might have
planned on sending their pull request a bit later. It's ok. Grovel a
bit, and explain what's up, and you can almost certainly guilt me into
Also, maybe I just missed something due to jetlag (hmm. yes, let's
call it "jetlag", that sounds so much better than "core incompetence
and bad planning"), so if you feel unfairly overlooked, send me a note
explaining how I've unfairly wronged you.
There is also at least one pull request that I am hoping to get asap
and planning on still pulling, ie I'm very much still hoping to get
overlayfs finally merged. But there were a few last-minute questions
from Al. Assuming that all works out, that's an expected late pull.
Not worth holding up the rc1 release for one known straggler, though.
So there you have it. The merge window is closed, but with room for
excuses and possible missed requests. As usual, the shortlog is much
too big to post (core stats: roughly 74% drivers, 10% architecture
updates, the rest networking, filesystems, core kernel, documentation,
include files, tool updates...), and the appended is my "mergelog"
which as usual credits the people I pulled from, which is not at all
necessarily the same as the people writing the code.
Go forth and test,